Documentation of lin_remove_NaN_defunct

Global Index (short | long) | Local contents | Local Index (short | long)

Function Synopsis

y = lin_remove_NaN(x,xtim,show)

Help text

lin_remove_NaN:  linearly remove a time series from data

   Y = lin_remove(Xdat, Xtim) removes the best linear fit of Xtim to
   each column of Xdat.  If Xdat is N-dimensional, then it is 
   assumed that the time series Xtim will be removed from the first
   dimension of Xdat.

   Y = lin_remove(Xdat) assumes Xtim is evenly spaced, so the linear
   trend is removed.

Cross-Reference Information

Listing of function lin_remove_NaN_defunct

function y = lin_remove_NaN(x,xtim,show)

sz = size(x); ndim = length(sz);
if (ndim == 2) & (sz(1) == 1); x = x(:); end;
sz = size(x); ndim = length(sz);

if nargin < 2; xtim = [1:sz(1)]/sz(1); end;

if nargin < 3; show = 0; end;

if (size(xtim, 1))==1; xtim=xtim(:); end;

if size(xtim, 1)~=sz(1);
  error('Xtim must have the same length as the first dimension of Xdat');
end

%  Reshape x if necessary, assuming the dimension to be 
%  detrended is the first

if ndim > 2;
  x = reshape(x, sz(1), prod(sz(2:ndim)));
end

%  Remove means from data and time series
N = size(x, 1);
xtim = xtim - ones(N, 1)*mean2(xtim);
x = x - ones(N, 1)*mean2(x);

%  Remove Regression

[N, m] = size(x);
[NN, mm] = size(xtim);

y = repmat(NaN, [N m]);

for i = 1:m;
    if show;
        disp(['Iteration:  ' num2str(i)]);
    end
    kp = find(~isnan(x(:,i)));
    for j = 1:mm
        kp = intersect(kp, find(~isnan(xtim(:, j))));
    end
    y(kp,i) = x(kp,i) - xtim(kp,:)*(xtim(kp,:)\x(kp,i));
end

%  Reshape output so it is the same dimension as input

if ndim > 2;
  y = reshape(y, sz);
end